The incidence and risk factors of coronary steal after ipsilateral AVF in patients with a coronary artery bypass graft

Abstract

Objective

Ipsilateral arteriovenous fistula (AVF) may cause symptoms of coronary steal in patients who had undergone coronary artery bypass graft (CABG) using internal thoracic artery. The purpose of this study was to evaluate the adverse effects of ipsilateral AVF to CABG, including the incidence of coronary steal, and to analyze the risk factors for coronary steal.

Methods

Between 2000 and 2013, a total of 25 patients undergoing hemodialysis via upper extremity AVF, ipsilateral to the preexisting CABG, were reviewed retrospectively. Clinical assessment related to coronary steal, echocardiography before AVF, and coronary angiography after symptoms were assessed. The definition of coronary steal was the new development of one or more of the following symptoms within 12 weeks after AVF creation: chest pain, chest discomfort, and dyspnea.

Results

Three patients were clinically diagnosed as coronary steal. Left ventricular ejection fraction (LVEF) was statistically lower in coronary steal group compared to no-steal group (41.7% vs. 50.9%; p = 0.036). Patients with coronary steal were older at the age of CABG surgery and showed a higher incidence of regional wall motion abnormality.

Conclusions

Coronary steal after ipsilateral AVF creation in patients with CABG using in situ internal thoracic artery (ITA) developed in 12%, which can be related to low LVEF. In patients with CABG requiring AVF for hemodialysis, the ipsilateral fistula to the grafted ITA should be carefully selected and performed.

Keywords

Arteriovenous fistula Coronary artery bypass graft Hemodialysis Internal thoracic artery

Introduction

Creation of ipsilateral upper limb arteriovenous hemodialysis (HD) access could cause coronary steal in patients with coronary artery bypass graft surgery (CABG) using in situ internal thoracic artery (ITA) (1, 2). Some reports showed different blood flow volume between ipsilateral and contralateral ITA during the use of arteriovenous fistula (AVF) for dialysis. In addition, angiography revealed no antegrade flow to ipsilateral ITA in some cases (3). This hemodynamic phenomenon can be explained by Venturi effect, in which flow to AVF has lower resistance than to ITA-coronary anastomosis (4). As a result, flow reduction to ITA and even reversal flow of ITA to distal subclavian artery could occur after creation of ipsilateral AVF in patients with pre-existing CABG.

Development of coronary steal may be associated with increased risk of hypokinesia, cardiac infarction and lower long-term survival in HD-dependent patients after CABG (5). On the other hand, other studies did not detect evidence of a coronary steal in HD patients with ipsilateral AVF (6). In addition, several recent studies demonstrated that ipsilateral AVF did not affect the result of cardiac event and long-term outcome (7).

There are no data on the incidence of coronary steal after ipsilateral AVF creation in patients with CABG using ITA. Considering the risk of kidney injury during CABG procedure, selection of AVF location after CABG could be an important issue to a vascular surgeon (8).

The purpose of this study was to investigate the incidence of coronary steal after ipsilateral upper arm AVF creation in patients who underwent CABG and to analyze the risk factors for coronary steal. In addition, we provided long-term outcome in patients with ipsilateral AVF.

Methods

The Institutional Review Board (IRB) approved this retrospective study, and the approval included a waiver of informed consent.

Between 2000 and 2013, a total of 2212 patients underwent CABG at our institution. End-stage renal disease (ESRD) was newly developed in 67 patients, who underwent arteriovenous HD access creation. Among them, 25 underwent vascular access creation (native fistula or synthetic graft) in the ipsilateral upper limbs, and were enrolled in this study (Fig. 1).

Fig. 1

A flow diagram of patient selection and allocation. CABG = coronary artery bypass grafting; AVF = arteriovenous fistula; ITA = internal thoracic artery; LITA = left internal thoracic artery; RITA = right internal thoracic artery; BITA = bilateral internal thoracic graft.

Data collection and follow-up

The clinical data for each patient included the following: sex, ages at CABG and AVF, hypertension, diabetes mellitus, follow-up period, CABG type, and fistula type. Data on left ventricular ejection fraction (LVEF) and regional wall motion abnormality (RWMA) by echocardiography and perfusion reversibility by myocardial scan were obtained during the period between after CABG and before AVF creation.

CABG procedure

The basic surgical procedures and principles of CABG have been previously described (9, 10). The left ITA was used in all study patients. It was harvested in skeletonized fashion and used as an in situ graft whenever possible (n = 23; the ITA was the only inflow source in 18 patients, both in situ ITAs were used in 3, and both ITAs and in situ right gastroepiploic artery [RGEA] were used as inflow sources in 2). In the other 2 patients, the right ITA was used as an inflow source because of the combined pathology of the left subclavian artery. Other conduits used for composite grafting were right ITA (n = 9), saphenous vein (n = 6), and RGEA (n = 4).

AVF site selection

In our center protocol, creation of contralateral AVFs is strongly advocated whenever possible. However, ipsilateral AVFs were performed in some cases due to lack of access vein, patient preference or miscommunication.

The clinical definition of coronary steal related with AVF

Coronary steal was defined clinically as the new development of one or more of the following symptoms within 12 weeks after AVF creation; chest pain, chest discomfort, or dyspnea, especially aggravated during HD. In addition, there was no other possible cause such as significant subclavian stenosis or acute myocardial infarction.

Statistical analysis

Continuous variables were expressed as means ± standard deviations and compared for differences between two groups with the Mann-Whitney U test. Categorical variables were expressed as percentages and analyzed using Fisher's exact test.

Cumulative death and major adverse cardiac events were calculated using the Kaplan-Meier method. All statistical analyses were performed with SPSS Statics version 18 (IBM Corporation, Armonk, NY). The results were evaluated at the p<0.05 significance level.

Results

Demographics

We identified 25 HD patients via ipsilateral upper limb AVF after CABG during the study period. Figure 1 showed a flow diagram of patient selection. Among them, 18 patients (72%) had in situ left internal thoracic artery (LITA) CABG and 5 patients had in situ bilateral internal thoracic artery (BITA) CABG.

The baseline demographic and clinical data of the patients are presented in Table I. Mean follow-up duration was 85.4 ± 54.6 months (median 97 months, range 4-174).

Table I

Baseline characteristics of patients undergoing ipsilateral AVF after pre-existing CABG in the 2000-2013 (n = 25)

Variables
Sex (men:women)	16 (64%): 9 (36%)
Age at CABG (y)	63.9 ± 9.2
Age at AVF (y)	68.0 ± 9.1
Hypertension	21 (84%)
Diabetes mellitus	19 (76.0%)
Follow-up (mo)	85.4 ± 54.6 (97 median, range 4-174)
CABG type
LITA	18 (72%)
RITA	2 (8%)
BITA	5 (20%)
Fistula type
RC AVF	10 (40%)
BC AVF	10 (40%)
AVG (forearm loop)	4 (16%)
Forearm basilic vein transposition AVF	1 (4%)
Symptom^a	9 (36%)

Chest pain, chest discomfort, or dyspnea confirmed by cardiac surgeon, nephrologist, or cardiologist.

CABG = coronary artery bypass grafting; AVF = arteriovenous fistula; LITA = left internal thoracic artery; RITA = right internal thoracic artery; BITA = bilateral internal thoracic graft; RC = radio-cephalic; BC = brachio-cephalic; AVG = arteriovenous graft.

Coronary steal related to AVF (n = 3)

Three patients were diagnosed as coronary steal by our definition (Tab. II). Patients in coronary steal group were older at CABG surgery and showed higher incidence of RWMA than those in no-steal group, without significance. LVEF was statistically lower in coronary steal group compared to no-steal group (41.7% vs. 50.9%; p = 0.036). Coronary steal did not occur in BITA grafting and right internal thoracic graft (RITA). No statistical differences were observed between groups for CABG type and fistula type.

Table II

Comparison of variables according to coronary steal

Variables	Steal (+) n = 3	Steal (−) n = 22	p value
Age at CABG (y)	69.7 ± 11.0	63.1 ± 9.0	0.206
Age at AVF (y)	70.0 ± 10.4	67.7 ± 9.1	0.723
LVEF (mean)^a	41.7%	50.9%	0.036
RWMA^a	3 (100%)	10 (45.5%)	0.220
Reversible perfusion defect^b	1 (33.3%)	8 (36.4%)	0.474
CABG type			1.000
LITA	3 (100%)	15 (68.2%)
RITA	0	2 (9.1%)
BITA	0	5 (22.7%)
Fistula type^c			0.565
Forearm	2 (66.7%)	9 (40.9%)
Upper-arm	1 (33.3%)	13 (59.1%)

On echocardiography before AVF.

On myocardial scan before AVF.

Forearm: RC + basilica vein transposition AVF; Upper-arm: BC + forearm loop AVG.

CABG = coronary artery bypass grafting; AVF = arteriovenous fistula; RWMA = regional wall motion abnormality; LVEF = left ventricular ejection fraction; LITA = left internal thoracic artery; RITA = right internal thoracic artery; BITA = bilateral internal thoracic graft; RC = radio-cephalic; BC = brachio-cephalic; AVG = arteriovenous graft.

The three patients diagnosed with coronary steal underwent AVF closure and switched to a contralateral side fistula. Before performing AVF closure, all of the patients underwent coronary angiography, which showed patent CABG without other abnormal findings. Among them, two patients improved in symptoms and one patient was lost to follow-up.

Long-term outcomes in the patients with ipsilateral AVF

A total of 22 patients received HD using ipsilateral AVF. Median follow-up was 90.9 ± 54.2 months (interquartile range, 42.5-135.8 months). Four patients died during the follow-up for meningitis, pneumonia, congested heart failure and ruptured abdominal aortic aneurysm. Late major adverse cardiac events, defined as death, myocardial infarction and the need for repeat revascularization by redo-CABG or repeat percutaneous intervention, occurred in 36.4% (n = 8) patients overall (Fig. 2).

Fig. 2

Kaplan-Meier analysis shows time until death and major adverse cardiac events (MACE) in patients with ipsilateral arteriovenous fistula (AVF) (n = 22).

During the whole follow-up period, overall six patients complained of chest discomfort. The average interval to chest discomfort symptoms was 824.4 days. The causes of the symptoms were right coronary artery stenosis confirmed by coronary angiography (n = 3), pulmonary edema (n = 1), and unknown (n = 2).

Discussion

After creating ipsilateral AVF in upper limb for HD in patients with a pre-existing CABG, three patients (12%) were clinically diagnosed as coronary steal syndrome related to AVF. All of them were treated with AVF closure and contralateral AVFs were made simultaneously. LVEF before AVF was statistically lower in coronary steal group and they tended to have RMWA.

Incidence and risk factors of coronary steal after ipsilateral AVF are not well established. Coronary steal can be diagnosed by clinical symptoms, electrocardiographic changes, Doppler ultrasound and angiography (1, 2, 5). Patients with decreased flow in grafted ITA could experience hypokinesia, ischemic electrocardiogram changes and angina. However, not all patients with coronary steal after ipsilateral AVF would experience symptoms (3). Therefore, the incidence of coronary steal defined by clinical symptoms may lead to under- or overestimation.

One important issue is whether asymptomatic coronary steal has an effect on cardiac events and long-term morbidity and mortality (11). In our study, mortality was 18.2% during the follow-up period, but possible relationship with coronary steal was found in one patient, who died from complications of congestive heart failure. Takami et al (7) also reported similar results on long-term cardiac event or mortality after ipsilateral AVF/CABG. They demonstrated that cardiac event-free rates (ipsilateral 74% vs. contralateral 68% at 5 years) and overall survival (ipsilateral 58% vs. contralateral 65% at 5 years) were similar between the groups.

Distance from ITA orifice to AVF anastomosis site could affect occurrence of coronary steal (12, 13). However, our data showed no statistically significant differences between upper-arm vs. forearm AVF. Likewise, a prospective study did not demonstrate any significant hemodynamic differences between radial and brachial AVF on color Doppler ultrasonography (14). Blood flow volume through the AVF might affect occurrence of coronary steal rather than its location.

Coronary-subclavian steal syndrome (CSSS) can occur in as many as 4.5% patients after CABG (15). The possible causes of CSSS are proximal stenosis of the subclavian artery, diffuse coronary artery disease with poor runoff and ipsilateral AVF (16, 17). In our study, three patients complained of dyspnea, chest pain or discomfort shortly after creation of ipsilateral AVF. Most patients would be tolerable because of cardiac compensation after AVF creation (18, 19). Therefore, ITA flow steal may occur in patients with high flow volume of 2000 mL/min through AVF and uncompensated cardiac output (6, 20).

We still prefer to create contralateral AVFs in patients with CABG. We observed no flow or delayed flow in unharvested ipsilateral ITA at pre-CABG angiography in patients with upper limb AVF. Kato et al also reported even retrograde flow in an unharvested ITA at angiography (3, 21). The principle of contralateral AVF creation after CABG has been applied due to the concern for the risk of coronary steal phenomenon. However, in some patients with limited vein, creation of ipsilateral AVF would be inevitable. In such cases, additional studies would be required to elucidate benefit to use ipsilateral upper arm vein compared with lower extremity AVF.

In such situations, also predicting the potential risk factors of grafted ITA steal is an important issue. Lower cardiac function could be a predictor of coronary steal after ipsilateral AVF. A study on the relationship between cardiac function and coronary steal is needed.

This study has several limitations. The diagnosis of coronary steal was made clinically, without imaging evaluation by Doppler or angiography. Therefore, it is likely that the incidence of coronary steal was underestimated. There could be selection bias as ipsilateral AVF because of our policy on AVF site selection. The statistical calculations for the relationship between coronary steal and possible risk factors had low power because of the small sample size. Therefore, we could only detect a relationship with the LVEF.

Conclusions

Clinical coronary steal after ipsilateral AVF creation in patients with CABG using in situ ITA, developed in three patients (12%), which could be related to low LVEF. These findings can be used to guide inevitably creation of ipsilateral AVF in patients with pre-existing CABG. In addition, further multicenter registries such as web-based data collection platforms are needed to get a better understanding of coronary steal after dialysis access using ITA.

Footnotes

Financial support: No grants or funding have been received for this study.

Conflict of interest: None of the authors has financial interest related to this study to disclose.

References

Kinoshita

Asai

Ishigaki

Steal from skeletonized internal thoracic artery graft during hemodialysis after coronary artery bypass grafting.

Heart Surg Forum. 2010;13 (4):E254–E256.

Gaudino

Serricchio

Luciani

et al.

Risks of using internal thoracic artery grafts in patients in chronic hemodialysis via upper extremity arteriovenous fistula.

Circulation. 2003;107 (21):2653–2655.

Rahbar

McGee

Birdas

Muluk

Magovern

Maher

Upper extremity arteriovenous fistulas induce modest hemodynamic effect on the in situ internal thoracic artery.

Ann Thorac Surg. 2006;81 (1):145–147.

Ritter

Ringelstein

The Venturi effect and cerebrovascular ultrasound.

Cerebrovasc Dis. 2002;14 (2):98–104.

Kobayashi

Changing strategy for aortic stenosis with coronary artery disease by transcatheter aortic valve implantation.

Gen Thorac Cardiovasc Surg. 2013;61 (12):663–668.

Tokuda

Song

Internal thoracic artery grafts and upper extremity arteriovenous fistula.

Ann Thorac Surg. 2007;84 (6):2138.

Takami

Tajima

Kato

et al.

Effects of the side of arteriovenous fistula on outcomes after coronary artery bypass surgery in hemodialysis-dependent patients.

J Thorac Cardiovasc Surg. 2014;147 (2):619–624.

Rydén

Sartipy

Evans

Holzmann

Acute kidney injury after coronary artery bypass grafting and long-term risk of end-stage renal disease.

Circulation. 2014;130 (23):2005–2011.

Kim

Cho

Chang

Lim

Ham

Kim

Bilateral skeletonized internal thoracic artery graftings in off-pump coronary artery bypass: early result of Y versus in situ grafts.

Ann Thorac Surg. 2002;74 (4):S1371–S1376.

10.

Hwang

Kim

Complete Revascularization of the Three-Vessel Territories Using a Left Internal Thoracic Artery Composite Graft.

Ann Thorac Surg. 2015;100 (1):59–66.

11.

Feldman

Tkacheva

Efrati

et al.

Effect of arteriovenous hemodialysis shunt location on cardiac events in patients having coronary artery bypass graft using an internal thoracic artery.

Ther Apher Dial. 2014;18 (5):450–454.

12.

Baciewicz

Jr.

Selection of dialysis patients for bilateral internal thoracic artery grafts.

Ann Thorac Surg. 2007;84 (5):1797.

13.

Baciewicz

Jr Harris

Is steal present in hemodialysis patients with ipsilateral internal mammary graft?

Ann Thorac Surg. 2002;73 (1):349–350.

14.

Coskun

Colkesen

Altay

et al.

Hemodynamic effects of left upper extremity arteriovenous fistula on ipsilateral internal mammary coronary artery bypass graft.

Thorac Cardiovasc Surg. 2013;61 (8):663–667.

15.

Sintek

Coverstone

Singh

Coronary subclavian steal syndrome.

Curr Opin Cardiol. 2014;29 (6):506–513.

16.

Elian

Gerniak

Guetta

et al.

Subclavian coronary steal syndrome: an obligatory common fate between subclavian artery, internal mammary graft and coronary circulation.

Cardiology. 2002;97 (4):175–179.

17.

Rossum

Steel

Hartshorne

Evaluation of coronary subclavian steal syndrome using sestamibi imaging and duplex scanning with observed vertebral subclavian steal.

Clin Cardiol. 2000;23 (3):226–229.

18.

Frank

Wang

Lammerant

Miller

Wegria

An experimental study of the immediate hemodynamic adjustments to acute arteriovenous fistulae of various sizes.

J Clin Invest. 1955;34 (5):722–731.

19.

Gaudino

Serricchio

Tondi

et al.

Non-invasive evaluation of mammary artery flow reserve and adequacy to increased myocardial oxygen demand.

Eur J Cardiothorac Surg. 1998;13 (4): 404–409.

20.

Crowley

Butterly

Peter

Schwab

Coronary steal from a left internal mammary artery coronary bypass graft by a left upper extremity arteriovenous hemodialysis fistula.

Am J Kidney Dis. 2002;40 (4):852–855.

21.

Kato

Ikawa

Hayashi

Yokoyama

Internal mammary artery steal in a dialysis patient.

Ann Thorac Surg. 2003;75 (1):270–271.